Fedora clean up process seems to be seriously broken...

[Michael Schwendt]

On Tue, 22 Nov 2011 00:09:36 +0100, RH (Reindl) wrote:

> 
> 
> Am 21.11.2011 23:50, schrieb Michael Schwendt:
> > On Mon, 21 Nov 2011 22:58:50 +0100, RH (Reindl) wrote:
> > 
> >> +1
> >>
> >> nothing is more frustrating for users as ignored bugreports reintroduced from
> >> release to relase while th eonly response is from bugzapper about EOL of the
> >> release
> > 
> > Well, that's not the same problem as this thread is about.
> > 
> > There a very active packagers (and developers who also do packaging tasks)
> > who don't respond to [all] tickets due to various reasons. Some of the
> > reasons are valid. Become a package maintainer yourself, Harald, before
> > you judge about them all.
> 
> i do not judge!
> 
> in my opinion there is no reason to not respond
> respond can be negative with a short "why"
> well, this would be much more helpful as bugzapper-mails

That's even another topic. ;)

Those automated bugzapper mails come *much* too late. They are a poor
attempt at cleaning up old cruft in bugzilla. They don't handle the case
of bug reporters refreshing tickets again and again in response to the
automated NEEDINFO query *without* any human being ever deciding to spend
time on the ticket.

However, for _some_ components of the distribution, if you want human
beings to handle the incoming bugzilla traffic, these would need to be
real bugzilla monkeys. Some components receive hundreds (if not thousands)
tickets per dist lifetime. In Fedora bugzilla. In addition to the upstream
bug tracker.

> it is a hughe difference if you give no feedback to anyone
> who took the time to make a bugreport or ignore it

Given the amount of bz traffic for some pkgs, it isn't easy to even try to
respond to them all. Especially if the devs are active in the upstream
ticket system, too. There is also no guarantee that the bug reporter will
be helpful beyond the initial report. Some reporters simply don't respond
anymore either. As I say it, ABRT can be both a blessing and a curse. ABRT
makes it much easier for arbitrary users to dump something into bugzilla,
under the assumption that the submitted backtrace is enough, even if the
problem is not reproducible.
 
> if the respponse is "sorry no time yet" is would be much
> morehelpful than no response at all

*That* should be an automated response. Including the request to consider
filing a bug upstream. An opt-in service for packagers to enable it for
their packages.

All not trivial, however, as _somebody_ will need to decide whether a bug
is specific to Fedora or whether it's a general bug in the source code.

At first you may be satisfied by a "sorry no time yet" response, but
for how long? Eventually you would demand a fix to be delivered, or
another response, or some kind of promise to spend time on the issue.

> we all know that nothing will be perfect now or in future but
> give users the feeling that they are not ignored is a high value

Are they? Not in general. We need to find out the _where_ and _why_.

It's not unusual for your own issues to become your pet peeves.
Instead of waiting for others to work on fixes, try to find an area
where to contribute.